How to make a Softstarter and why it is sometimes mandatory to use!

[@ParenteGrande]

During my education we were taught about this. We were taught it was called a Startup Power Spike and most electronics are designed/engineered to reduce this but normally they remove the circuits in production to cut costs. I haven't done any of this in over 20 years but I still have my textbooks and notes. I miss tinkering with circuits. Thanks for the memories and the great work.

[@Metroid1890]

Your content always saves me and my teammates right before working with some power electronics, you're a lifesaver

[@greatscottlab]

You're welcome :-)

[@KerbalLauncher]

Make sure you employ high side switching.

[@Metroid1890]

@@KerbalLauncher By High switching you mean actually connecting the mosfet before the relay?, also I wonder if this could work at 24V aswel, since the battery we are using has a reference there

[@KerbalLauncher]

@@Metroid1890 No, I mean don't even use a relay. See the datasheet for this part: www.analog.com/media/en/technical-documentation/data-sheets/LTC7004.pdf Page 12 details inrush current limiting methods.

[@Metroid1890]

@oH well,lord! even so it seems pretty unreliable, I like the way he explains it by applying raw theory in simple circuits

[@akshaykumargautam5289]

"stay creative " is the best dialogue you say. Keep up the awesome work

[@shaileshkullal11]

Why are u so gentle and genuine person? BDW i had this problem 4yrs back, while designing pwm controller for big dc motor drive. I've just use few passive components & added some start delay in microcontroller code, which was initiating pwms to drive H bridge.

[@monrax]

GreatScott: How to make a softstarter Also GreatScott: That's a subject for another video

[@rogercoulombe3613]

Motor controllers are typically another beast entirely to the circuit shown in this video. It makes sense to separate them

[@santi308]

GreatScott: JLC Pcb... blah blah blah. Also GreatScott: *Do circuits with Perfboard*

[@aged_mozzarella]

How to control motors and softstarters for them. Pls

[@jaycool5978]

@@aged_mozzarella depends how you want to control them and the power required for them. Often 1 uses a motor driver. H bridge is a configuration of Mosfets. Another awesome motor driver is a Sabertooth 2. Look into them see if they're what your looking for. Often they can be controlled at logic level around 5 volts, and many configurations accepts PWM, which is very useful for varied RPM applications. Some can be controlled from a RC receiver.

[@aged_mozzarella]

@@jaycool5978 Thanks, but the thing I always struggle with is the controlling of the h bridge or other configurations. I am not very sure about the relation between frequency and duty cycle. I couldn't really find anything on the internet that'd explain this

[@Minimalist11Guy]

Thank you so much for the well-presented KZbin. I need to know a lot about soft starting as I am using four 3-kilowatt quartz infrared heaters, that can be switched with a remote controller, to one, two, or three kilowatts. I think that they all need a 16 Amp type C, circuit breaker. The standard circuit breaker trips immediately. On further research, I find that with a soft start, the extremely expensive heating tubes will last 30% longer. Now back to my experimenting!

[@pau1976]

You can also use a coil in series and a protection circuit to discharge it safely ( diode and small resistor ). It's quite an easy and simple solution. Besides, you can also add in parallel a capacitor to with a resistor move all the high frequencies back to the source. These are pretty simple strategies most power supplies use.

[@nick86i]

I think you need a pretty big coil made with very thick wire that can handle a few tens of amps required by such an inverter and maybe it's not suitable for an inverter

[@HPD1171]

the circuit i have been using for a while is really simple. it consists of a P-channel FET a small capacitor and a couple resistors. the P-channel fet is wired as the high side switch with the source connected to the input side and the drain connected to the output. a small capacitor is placed between the output and the gate with a 100k resistor accross it to keep it discharged and a 10k resistor from gate to ground. when the device is powered the transistor is normally off since the capacitor which is place between the gate and drain starts a zero volts the rising output voltage acts as a negative feedback loop limiting the rate at which the output can switch on and once the capacitor is fully charged by the 10k resistor to ground the transistor will be fully enhanced and the output will be on. the rise time of the output can be adjusted by changing the feedback capacitor and resistor values. a zener can also be placed from gate to source to limit the voltage across it in cases where the supply voltage exceeds the gate - source voltage rating of the transistor. when the device is powered off the resistor across the capacitor will discharge it in a few ms so it is ready for the next power on. I first saw this when at university and thought it was a very elegant design that does not need and relays or complicated timing circuitry but can't remember what it is called other then P-MOSFET current limiter. but when checked on the oscilloscope it always produces a very nice perfect ramp up to the final target voltage which is the ideal wave form shape for charging a capacitor though there is a small jump from zero volts up to whatever the threshold voltage of the mosfet is so you should pick a fet with a low threshold to reduce the initial spike but other then that it works great.

[@gladosai4354]

Human, I understood everything you said. Because it is science!!!! You are my favorite youtuber human!!!

[@hashvolting]

Send wheatley to my home made ranch

[@gladosai4354]

@@hashvolting ok then... But when that idiot is on his way... He is gonna be like a meteor

[@joopterwijn]

Are there other kinds of KZbinrs then “human”... 🤪🥴

[@Electricmemerat]

where's the moron?

[@hashvolting]

@@joopterwijn did you read his name?

[@r1273m]

I told the checkout girl at my local supermarket that if they had a softstart on the conveyor my milk would not fall over every time the belt moved up. She gave me a very strange look and told me to lie the milk down...…. Ah well, good try!

[@Miata822]

Good video on a common issue. I made a similar circuit several years ago for use on a 70A 24VDC set of motors I had to start with a 100A power supply. I really had to let the motors get well started before applying full power. I used a large high power 3ohm resistor and an electronic delay-on-make relay switching a larger relay made to withstand constant use at high current and temperature. It worked well. I've seen a guy build a similar device to start a helicopter jet engine. On mine, since this was used on a commercial project, I had to use modular components to allow for field service. I also included a one-shot temperature limit switch and yet another relay to ensure that if the time delay relay failed the big resistor couldn't overheat and cause a "thermal event" (what we called fires back in the day). I've also used NTC thermistors on things like 10-15A blowers that weren't a threat to the power supply but provided less of a challenge to my customers' often overtaxed circuit breakers. You want to choose thermistors carefully and use them fairly close to their rated power so that you get minimal voltage drop over them in steady state use. I look forward to your next soft start video.

[@Miata822]

VIPIN LADWAL if you enjoy these videos and learn thing from them you could become a patron and not always be 5 days Late :)

[@007hansen]

@@LAZYLONER- and ​@Bill Kerr, this abo theme certainly is the next era of capitalism. Call me names but I prefer the single deposit.

[@victornpb]

One time I used a toaster in series as a current shunt to soft start a compressor

[@Miata822]

@@victornpb Ha! Perfect.

[@victornpb]

Not everyone has big power resistors at disposal, except they do, 900W toaster worked perfectly!

[@jstagzsr]

ive been on a kick lately of taking apart every piece of electronic equipment i have laying around my house not in use and mining it for all its little components. Watching your videos give me so many ideas for what to do now with this box of random parts.

[@Sousanators]

Slight error with your RC time constant: The resistance seen by the capacitor is the thevenin equivalent, not the series charging resistor. Your time constant should be 0.5s, verified in simulation

[@Nic7320]

He overestimated it, but that's better than underestimating it. Still, 0.5s is still much longer than the RC delay needs to be, since the input transient he measured is only 15 ms and there's more delay starting up the power converter. But he also made another *simplifying* assumption, when he used the capacitor ESR to estimate the input transient. The MPPT output impedance, and the wiring resistance, and the capacitor's ESR all limit the current charging the input caps.

[@keithatkins9394]

@@Nic7320 Smart Alec!

[@fmphotooffice5513]

@ 3:00 ONE HAND! It doesn't matter you know the clip is plastic. As standard procedure, only one hand when on any powered circuit, **especially** when your left hand is on a metal chassis. The dead engineers wish to tell you they also "thought" it was ok. (Just trying to keep you with us as long as possible.) Regards.

[@jakerezac9088]

I love seeing these smaller but still very useful circuit projects. This is gonna be super helpful on my next project!

[@pranavasthanaDelhi]

I just completed this circuit and it works like a breeze. I was trying to run a dc motor using atx power supply boosted by a boost converter (12v to 24 v). Initially, due to inrush current, atx power supply was getting shut down but now it works well. Thanks Scott!

[@jorno1994]

stick with the NTC but use it with the bypass relay. it's exactly what high end computer PSUs do nowadays to get over 90% efficiency ratings.

[@devrim-oguz]

It would still have the problem of remaining hot.

[@1Hippo]

@@devrim-oguz Should not be a problem, because the NTC already cools down when it is still running, but the bypass is engaged.

[@Nathan0A]

@@devrim-oguz depending on how fast the caps discharge, perhaps while the NTC is still hot, the caps are still charged and inrush is limited this way? It seems like the best way to do this would be to have a circuit that isolates the caps, soft charges them for a second, and then uses a multi throw relay to disconnect the soft start circuit from the caps, connecting caps to main circuit, and connecting power to main circuit at the same time

[@soupflood]

That would still be defeated when the appliance is connected and disconnected in a matter of seconds, leaving no time for the thermistor to cool down.

[@aarriikknn33ll]

Ali Devrim OGUZ yeah no shit, there will be some energy loss but it will get less hot.

[@MaxMakerChannel]

I had this problem when I built my electric hydrofoil. The ESC had a huge current spike. I solved it with two relays that bypasses the current through a big resistor for one second before the main forklift relay is activated.

[@yangyan6645]

Will an inductor in serial also solve the problem? An inductor will reduce the current surge.

[@teardowndan5364]

The problem with limiting inrush current using inductors at low frequencies (DC in this case) is that inductor values tend to be insane. Typical inductors in everyday electronics range from 100nH to 100mH but here, we'd most likely be talking 100+ henries. Also, you'd then have to deal with safely discharging all the energy now stored in the series inrush-limiting inductor when you turn off or disconnect the load, otherwise something will get destroyed by the collapsing magnetic field dumping its energy into whatever weak spots it finds.

[@yangyan6645]

@@teardowndan5364 Thank you, your explanation is very clear!

[@TheRealKuif]

@@teardowndan5364 Wait, why would we need 100+ henries exactly? A few mH would be more than enough to limit the spike somewhat. A (physically) big inductor with a low resistance would be just fine. Simply connect it in series right at the output of the DC source and stick a diode from the negative to the positive terminals to deal with the collapsing field.

[@teardowndan5364]

@@TheRealKuif Right, I must have inverted something. Tens of mH at a handful of amps is still going to be a chunky core if you want to avoid saturation and a chunky winding.

[@glasser2819]

I guess what we are dealing with here is the inrush surge current to charge the inverter capacitors, right... was thinking about zero-crossing soft-starter but this is a DC circuit, right? then use a simple car light bulb in serie with the load: will glow until current gets null ie. caps charged! Use a 3 way an "ABC Switch" to switch input between bulb then direct. 👍

[@Glenn.Cooper]

I really love the way you deliver content. For someone coming at this with a reasonable technical background, you deliver very relevant information in a very easily understood way. If I could make one recommendation it would be this: Don't change a thing!

[@AKSnowLeopard]

Most people: This electrical device doesn't work, I'll toss it and get a new one. Electrical engineers: Hold my beer...I'll make it work.

[@Minitoys59]

because obviously we are always holding a beer ;)

[@caffeinatedinsanity2324]

Drunk science for the win!

[@JerryLee..]

Boom!

[@satibel]

- "The plug started sparking so I threw it" - just change the plug and the cord, dammit.

[@MoxxMix]

How about not putting unnecessary comments.

[@dnyaneshvarsalve]

I like the way everytime you write down and show us the practically calculated values of resistance, current and voltages

[@travispulley8899]

I enjoy these videos so much! Your presentation is consistently excellent and sufficiently detailed, and with very good production

[@SeanBZA]

Did a similar thing for a large variac, that would trip the main breaker at switch on. Just some 10W 12R resistors in parallel, with a thermal fuse attached to them, for failure protection. Then a simple capacitive voltage divider, driving a 24V relay after the bridge rectifier, with a 470uF 63V capacitor and a 24V 1W zener diode, along with a 33R 1W inrush limiting resistor for the capacitive divider. Relay would close contacts about 0.5 second after mains was applied, but enough time for the variac core to not saturate. All parts laying around, and worked well. Also added a voltmeter to the variac, as it had no scale, and made one out of an old meter movement, and as the variac can go to 360VAC, wanted a suppressed zero, so used a tap ( 60VAC) on the variac to derive a regulated 7V supply, using a LM723, and this was then connected to one side of the meter, with the other side getting a suitably half wave rectified output sample. Calibrated, remarked the meter and it works well.

[@fifaham]

You may use variable R2 and R3 to adjust the time delay for charging and discharging the delay circuit thus making it adapt to different circuits configurations. Good job.

[@lanreakerele4375]

GreatScott is such a good teacher for learners to learn about something important

[@lolkusus]

7:46 in this case time constant is actually C * (R2 parallel R3)

[@rjordans]

I'd probably add something for quick discharging of the 10uF capacitor when the input voltage is removed. Now this circuit has the same problem still as the NTC in that the switch-on time is reduced quite a bit when this capacitor isn't discharged yet. Though with the much larger delay compared to the pulse length you may indeed not notice that effect anymore in the current setup. Careful though if you have a MOSFET with a much lower threshold voltage

[@cornbread-glimpse-emblaze]

hahaha yes i also noticed this, seemed like such a hassle to get eventually the same results as the ntc lmao

[@MrSummitville]

Did you do the calculation to see how long it takes to discharge the 10uf capacitor ?

[@MrSummitville]

@@cornbread-glimpse-emblaze Did you do the calculation to see how long it takes to discharge the 10uf capacitor ?

[@cornbread-glimpse-emblaze]

@@MrSummitville tau = RC = 1 second to discharge to 27% of the initially applied voltage (see en.wikipedia.org/wiki/RC_time_constant)

[@architt4331]

I have absolutely no idea to what softstarter is, but I still watched your video as it is always very entertaining.

[@CoffeeConsumer643]

soft starter slowly increases the power input to the device that you connect it to its used with motors to control their start and make sure the machine wich has the motors doesnt malfunction one it is started up

[@spagamoto]

Good timing. I'm just about to put an NTC on this tube amp I built a while ago. No relay... not worried about a fraction of a watt, the thing burns dozens of watts just keeping its filaments warm :P The reason I'm doing this is not actually because my supply can't handle it, but because the tube filament lifetime is eroded more quickly if you hard start them... the filaments have a low resistance when cold and the resultant surge erodes their coating.

[@rallokkcaz]

Hey I noticed your videos have been absent from my feed for a very long time! Please keep making more content like this, it's super awesome as a hobby electronics nerd to see a master at work.

[@scottt5822]

After watching your video's I feel like making something really cool... but then I come back to the reality that I only know how to change the batteries in a torch. I love watching your video's. They inspire me.

[@AnotherBrokenToaster]

This is great, I like your basics videos more than large projects. But you made a mistake with the time constant, the equivalent resistance would be 50k because of the other resistor going to ground, which would give you a time constant of 0.5s

[@squareone2520]

Isn't that accounted for since his time constant was in relation to an applied voltage of 6V?

[@JotaFaD]

@@squareone2520 The time constant does not depend of the input voltage.

[@AnotherBrokenToaster]

@@squareone2520 It's kind of a trade-off, the voltage becomes 1/2 but the resistance is 1/2 also so it charges twice as fast. You can see at 8:38 that discharging takes twice as long as charging(because discharging uses only the 100k resistor connected to ground) and you can also kind of measure from the horizontal divisions(1s at 8:38) that the time constant is 0.5s when charging and 1s when discharging.

[@stask85]

Got exactly these problems currently. You explain the problem and solution VERY well. Thank you!

[@erygion]

I love the solar system stuff, when I have my own "Lab" I'm going to do my best to power it in this way. Great video!

[@gerardosandoval9163]

Man you're amazing, I spend part of my breakfast watching your videos, I find it clear and easy to understand thanks from a mechatronics engineering student.

[@acmefixer1]

There should be a 1N4004 diode across R1, with cathode band to +12V. This will immediately discharge C1 and 'reset' the circuit if the 12V drops to zero. Thanks for the great video. 👍

[@cdrive5757]

Across R1???

[@th34lch3m1st]

Speaking of coincidences, I was looking for an EMC/EMI filter right this days, because every time someone turn on a drill, a big fan or an air conditioner in my house, sometime those spikes turn off even my UPS-connected PC. I think an EMC/EMI filter would be a good subject for one of your videos. Protect digital devices from spikes that also comes from the main grid would be useful for everyone. Interesting video as always, thanks for sharing.

[@BladeforgerKLX]

I love it! If you can expand this to an A/C circuit, that would be great! When a refrigerator (for example) powered by an inverter hooked to backup -power-supply batteries causes the inverter to shut down, that's a real problem. Yes, a larger (and more expensive) inverter is one solution; however, some type of soft starter is a more elegant solution. THANKS!!!

[@FirstLast-vr7es]

Extremely useful! Thank you! I have been in need of a soft start circuit for tube amps. Lots of capacitance on the front end, and it can be a massive current surge at turn-on. I have a few 220uF 700V film caps, and they pull so much current on power-on that they have smoked some pretty beefy resistors that could easily have handled normal current draw. I can also use this for loudspeaker delay circuits. You explained it very well.

[@ancapftw9113]

Usually, your charger and inverter are both connected across the battery bank. You can also add a supercap to deal with fast, high amp loads like motor startup. Usually this would be how you deal with high amp draw. Not sure why it isn't working in your system. Is your inverter tied directly into your inverter without batteries?

[@Nic7320]

Victron MPPT controllers has over current protection on their DC output that doesn't like high initial surge currents. That particular 75V/15A solar controller only provides 15 Amps output anyways, not nearly enough for that 600 watt inverter at full load, which will draw about 60 Amps. So let's assume he using it as an example of a high input capacitance load for illustration purposes. I'm in a similar situation; I really like the diagnostic histogram panel the Victron MPPT controller provides on the bluetooth interface, so I put some of my low current items on the Solar DC output. But, like you point out, my 1200 watt Victron inverter has to go directly to the battery.

[@frr5004]

Hi there Scott, I've stumbled over this video mostly by mistake and it has prompted me to write down my own impressions on the topic. DC power supply systems with a beefy central PSU, or even a battery, have always been present in Telco and industrial process control settings, and in spite of that, there are hardly any dedicated DC soft-starters on the market, and the makers of "powered devices" mostly do not care very much either. Now with the rise of solar systems, the number of people hitting this obstacle must be growing again. Historically I've done some reading on the topic, and during the years on my techie job, I've been mostly frustrated about the "state of the art". Currently the most thorough DC soft-start and protection circuits that I've met are based on the TI TPS2492/2493 = originally a hot-swap controller IC, driving an external power MOSFET. Here's my humble webpage on the topic: support.fccps.cz/download/adv/frr/inrush/inrush.html

[@zolatanaffa87]

You say: "talking about electric engines.....but this will be subject for another video". Exactly what I wanted to ask you, perhaps using a circuit that excites the relay at the start, only for the time necessary, to avoid the continuous excitation of the coil, which could interrupt and generate malfunctions. thanks in advance for future projects and for sharing your knowledge

[@thecuriousone9342]

If the mppt is always on you could connect a bigger capacitor (more than or equivalent to appliance input value) directly across the output terminals by externally charging it first and parallel with mppt output, so when there is an inrush the capacitor will provide the inrush directly to the inrush consuming device long enough to charge the internal caps, no complex circuit and no power loss

[@dartysengineering7275]

Hi GreatScott, hope that you will very soon make a softstarter for ac motors because my lathe has a 10kW motor and trips the breaker every now and then and my dad doesn't really appreciate the flicker of the lights at the beginning. keep up these awesome videos!

[@millomweb]

Use a star-delta starter for the lathe.

[@dartysengineering7275]

@@millomweb point is it's a 230/400v 3 phase motor and it is already wired in star because every winding is only suited for 230v but if you have any solution i'll try

[@millomweb]

@@dartysengineering7275 I can't say you're making sense !!! 400(415)v 3ph motor and then say each winding only suited for 230v ? Do I detect you've connected a neutral wire to the windings ? If so, try putting a switch in the neutral - turn the neutral off for starting and on again once the motor's running with minimal load. I'll have to think about whether you're actually running the motor on full voltage as you have it like that !

[@ambaro0887]

I find these lessons very therapeutic and mind soothing..💆‍♂️

[@nicolasg3305]

can't u put a capacitor before the inverter switch to charge the internal capacitor off the inverter just enough to prevent huge currents spikes ?

[@awesomefacepalm]

Well the current surge was caused by caps so my unqualified guess is no. It would only add another source for big current surges

[@alexoja2918]

The capacitor would then act as a problem itself. It would have to be substantially large.

[@MusicBent]

Another theoretical solution would be a series (aka shunt) inductor before the inverter. However this has two problems. 1)The inducter would create an LC circuit which may cause instability, especially with any active or switching circuits (like an inverted), and 2) many circuits would require massive inductors in proportion to the capacetence and in-rush current.

[@MusicBent]

Another way to think about it: Capacitors resist sudden changes in voltage (by creating large current draws) Inductors resist sudden changes in current (by creating large voltages).

[@BirdyBenJedi]

Great Idea, I use this kind of automated system already since about 20 years for powering on transformers (fuse switched off sometimes) and yes, for PC's also. They have with their switching power supply also a lot of capacitance within their infeed circuit. Sometimes there are quite big sparks at the switch or even at the plug itself. One day in very past I even lighted up a fire in the socket while plugging in such a PC - cable. The socked was dirty, may have had also some invisible oil inside. The spark caused by that capacitance at the plug had been enough to light that oily dirty thing within the socket up and a great flame came out of the socket against me. That experiences caused me to solve that kind of risk.

[@lokeyyyy]

"I will see you next time" was bit energetic this time 😀

[@SirDella]

It sounded like someone was playing with the pitch knob xD

[@_FRANK_LIN]

But never seen

[@wavegreen]

autotune haha

[@lokeyyyy]

@@wavegreen hahaha

[@Wurstbam92]

Sounds like demolition Ranch 😂

[@DMC428]

Edit at the bottom! Hi Scott, I'm looking for a softstarter for my parallel push pull tube amp and was glad to see you publishing this video. My problem is, that my relais activates instantly, as soon as I apply power (I'm breadboarding with 5 volts DC, 500 mA). Even with a 470 uF cap. I'm using a finder relay (40.61.9.012.0000), 100 k resistors and your suggested n-channel mosfet. I have to drive that softstarter with 6,3 volts later, because that's the only low voltage source in my mono blocks and I don't want to put something else in it (the less stuff in there, the better). I tried to reduce the resistance on the caps kathode but nothing changed. Hi, I found my error. It was in the voltage divider section and the connection to the gate of the mosfet. Had it connected the wrong way.

[@aswinvr1912]

Hi GreatScott... Is it possible to remove the relay and use only the mosfet.

[@alexoja2918]

Yes, however all the current would then go through the mosfet and stress it, requiring power components and producing heat. The problem this presents is a greater one than the required relay current.

[@victornpb]

Yes

[@GoodElectronics]

It's possible, but in mosfets there is a voltage drop.

[@jaa185]

yes of course, but you need to drive mosfet tightly,say by opamp..

[@psialt9720]

Yes, but mosfet has voltage drop, and it will heat up

[@davida1hiwaaynet]

Thanks for this information! Variable frequency drives which have massive capacitor banks in them use a "precharge" resistor and a bypass relay similar to what you have. They, however, use their microcontroller to determine when the capacitor is charged enough to close the relay. In these VFD drives, the input rectifier will fail if this circuit is not in effect.

[@prashikkhare14]

You have completed 1.1 m subscriber still not uploaded the drone video

[@greatscottlab]

Maybe next year, or the year after that. Not sure.

[@SuperSilver301]

The quadcopter one

[@Cinemaaereo]

@@greatscottlab Maybe at the Linux's year. Right? :)

[@KangJangkrik]

@@greatscottlab Seems like u need a drink for a while

[@anycatdessai]

Don't ask him about the drone project. He doesn't like to be hurried. Let it complete fully.

[@tucker934]

Just built this for a WW2 dynamotor, and it works great, I was trying to find something to limit the current so that it’s not hard on a PSU if I ever get one for it, I will say your schematic is very well made!!! Thank you very much GreatScott!!!

[@aidanwansbrough7495]

This was really interesting, love how clearly you explain things!

[@vaseemmehrancp9372]

Unlike the other channels I am watching, I consider your videos like reference study materials 👍 Good works bro

[@EngineerNick]

I wouldn't have been able to design so many of my projects without this channel. Thanks for another great video :)

[@alejandrodelabarra2838]

Very tidy. I could never do your extremely tidy explanations. Congratulations!! Thanks for sharing your knowledge and experiences!!!

[@silverjohnson3163]

Could you use an inductor to initially slow the inrush current and then when the magnetic field is setup allow all the current?

[@frr5004]

To all those people here making comments about mains AC motor soft-starters: Mains-level voltages are dangerous, that's where DIY gets a dark tint. But to the point: years ago there used to be an analog chip called the U2008D, which you could use to steer a triac directly. A single triac = single phase. It was capable of soft-start and it did have some features for motor startup specifically (what with the phase shift between "V" and "I" and maybe trying to address some of the deeper magnetic stuff). That chip has been out of production for like a decade now. And even when that was available, you wouldn't be able to use that (alone) for 3-phase controller designs. Nowadays apparently all the commercial products are using microcontrollers, I believe there are even dedicated uC models for 3-phase motor control... which makes DIY a little far fetched. Some opto-couplers or IGBT's would not be that much of a problem, but coding the real-time phase controller for a miscellaneous MCU probably doesn't quite pay off to a hobbyist... Besides, starting a 3-phase mains AC motor by basic phase control is not very efficient. Optimally you should change the mains frequency too, and have feedback about current consumption / zero crossings, etc. This is what "variable frequency drives" are about... expensive and noisy bastards :-)

[@NIKFLICK707]

Can't we use inductor in series with appreciate value so that it will limit the inrush current and as soon as the voltage gets stable we will have no inductive reactance and hence no resistance/powerloss (except for resistance of that wire)???🤔🤔

[@greatscottlab]

That is why the inverter came with an inductor on the input. But the value would need to be very high to get decent results. That would be too expensive since the inductor needs to be able to handle a lot of current.

[@NIKFLICK707]

@@greatscottlab oooh thanks for replying 😍 😍

[@tunahankaratay1523]

@@greatscottlab Have you made calculations about coil properties? Since the pulse period is long, you could use an iron core of a ballast of a fluorescent lamp(15ms would equal to around 150 hertz in ac).

[@epiendless1128]

I believe the fact that a thermistor takes a while to cool down is why we are sometime advised, when "turning it off an on again", to leave the appliance off for a few 10s of seconds before turning on again. Some people were asking if an inductor would work. I found a 2mH inductor limits the current nicely, but a quick check at farnell suggested they're not easily available in high current ratings. An inductor will also limit the PSU's ability to handle power surges, and can overstess the circuit by oscillating with the capacitor. In simulation with a 12V source, I was able to get the capacitor voltage up to 22V on startup, before it settled back to 12V. A field applications engineer from Linear Technology put me wise to the dangers of suddenly connecting a PSU with an inductive lead - the PSU can momentarity see TWICE the supply voltage, and be destroyed.

[@vijayendirangiridharan6113]

Greatscott : "don't hesitate to check JLC PCB....for their fast production capacity and delivery times!!!" ALSO GREATSCOTT : "MAKES HIS OWN CIRCUIT BOARD.".....TOTALLY SAVAGE🤯🤯🤔

[@tunahankaratay1523]

@@vojtadeges720 tru

[@ZsomborZsombibi]

PCB is useful *after* a dev board proves the circuit really works. If it does, I always order a PCB for a reliable, small, long-life, mountable item.

[@tunahankaratay1523]

@@ZsomborZsombibi 2expensive4me

[@tunahankaratay1523]

@Chris Russell Yeah true. The biggest problem is shipping for me. The shipping is $20+ and I usually need only 1 board. But I love SMD soldering too and I hope I can get a proper etching setup at home.

[@artbyrobot1]

@@tunahankaratay1523 all you need is a laser printer, an iron, some transfer paper, and some white vinegar, salt, and hydrogen peroxide and some copper clad. not hard bro

[@shouldbewillbe2813]

Such a miniscule thing, but so important like matter of life and death of electronic devices and components !!!!! Thank you so much for the insight Scott !!!!!

[@guerrillaradio9953]

Would a large capacitor/inductor pair, with the inductor in series on the supply side of the capacitor, solve the problem almost as well? I'm not criticizing, I'm trying to learn. Thanks!

[@artbyrobot1]

was thinking the same thing but don't think the capacitor is needed as that is already in the inverter. Just a inductor in series would be needed I think?

[@Nic7320]

@@artbyrobot1 -- Sure, in theory it can work, but do some math and you'll find the inductor would be very large and very heavy. And then you'll have another problem unloading the inductive current fly back when you turn the MPPT DC off, and if you don't add a big diode to catch it you'll destroy other parts.

[@shroomy7199]

Another great video ty Scott. I could listen to you telling me how paint dries you would make it sound so interesting. Keep up the great work

[@Cheese_1337]

Usually inverter is connected straight to the battery.

[@beedslolkuntus2070]

Even I was thinking that!

[@Nic7320]

Agreed, but Victron smart MPPT controllers have a nice histogram feature going back 30 days that makes their switched DC output very useful diagnosing problems. My DC powered 'fridge was shutting off and I didn't know it until I discovered it on the history page.. I wish all their products, including their inverters, had this.

[@beedslolkuntus2070]

Nic Oooh

[@millomweb]

@@Nic7320 So what was the cause ? Not powered direct from battery ? ;)

[@Nic7320]

@@millomweb I'm running two refrigerators and the rest of my RV off of a 24 volt Tesla battery module. The cause of the shutdown was actually the fridge; it thinks it's supposed to shut off if the input voltage drops below 21.5 volts, which is designed for a 24V lead acid battery, but not the proper voltage cut-off for my Tesla LiIon battery. The Tesla 6S LiIon module voltage range is 25 volts down to 20 volts. I had assumed that since the fridge could run on either 24 volts or 12 volts, anything between them was okay. But it turns out the fridge has two shutdown voltages for either 12V or 24V operation. My solution was to put in a 24 to 13.5 volt DC-to-DC converter so the fridge always stays in the 12V operating mode, even if the Tesla battery goes down below 21.5 volts. Then below 20 volts, the MPPT controller is set to shut down the DC output to protect the Tesla battery. Maybe that's a long answer to your question. I do run my inverters directly from the battery, but smaller DC loads, like the fridge, I like to run off the MPPT controllers so I can see the histogram.

[@halbvoll1]

Good Job, I had a similar issue to handle with a big 1.2KW Transformer, my solution needs to be at the 230V side, so I build a simple capacitor dropper circuit that barely supports the current for the relay and added a capacitor in parallel to the Relais. I had to do this 2 times in series because of the big startup current of the transformer

[@id15807936]

Wouldn't it be better to just use a MOSFET to bypass the resistor instead of using a relay ?

[@cattflap1447]

+1 I thought that too, is there a reason why you can't do that ?

[@1Hippo]

Yeah thought the same, maybe because the resistance of a closed relay is lower than a FET, which has still a low RDSon? But with his low current (

[@marksawyer6242]

You could implement something simple with a P-Channel MOSFET but the Rds ON is relatively high, probably similar to a relay, or maybe worse than a relay. If you want minimum losses, you would need an N-Channel MOSFET, but that increases the complexity because you need a Gate drive voltage which exceeds your Source voltage by around 10 or 12 volts typically. That would require a charge pump circuit to generate the required drive voltage. That's a lot of extra components. As with everything in electronics, it's always a trade off between cost, efficiency, reliability, space, development time. Did I forget anything?

[@id15807936]

@@marksawyer6242 How is it even a problem to get 12v to drive a mosfet if there is a 12v source in this circuit? and yeah i meant N-channel mosfet being connected between the load and the minus of power source

[@marksawyer6242]

@@id15807936 You could theoretically do as you say and implement the inrush current limiting circuitry on the negative line to make it easier to drive the N-Channel MOSFET. However, i would never recommend it. Messing with your return path is generally bad practice. Consider what happens to the system when the inrush current goes through your resistor (before the MOSFET has switched on) the return line of your entire system will be at an elevated voltage, due to the potential created across the resistor when the inrush current goes through it. Then your MOSFET would switch on, and suddenly the entire system jumps back down. That may be ok if you are operating your system in isolation. But consider the effect that could have if you have multiple electronic systems connected together, and one of them shifts its return potential level relative to the other systems. If your computer PSU did that every time you switched your computer on it could easily cause damage to the peripherals that were plugged into it, due to potential differences creating ground loops.

[@xanataph]

I made a softstarter last year for a circular saw, to make sure there was no possibility it would pop an inverter when turned on. It runs the saw through a resistance (the ceramic element of an old fan heater - so it can handle a few watts lol) for a good 5-7 seconds to make sure that it's spun up before applying full power. It was a little trickier to design because rather than having the power switch ahead of the softstart unit, it had to react to the saw's own trigger switch. Now the real interesting effect though, is if I use it with small power tools such as a multi tool. These will not draw enough current to trigger the timer so the resistance stays permanently in circuit. However, this causes the tool's speed control to work when normally it wouldn't when powered from a modified square wave inverter.

[@ahmetzekiuslu3331]

Nice video. Don't forget the A class amplifier. :)

[@DIYBill]

Audiophile grade

[@DIYBill]

@GreatScott!!

[@michelfeinstein]

Years ago I wanted to make one like this, but with reverse-voltage protection as well and only solid state components, no mechanical relays, good job.

[@maciekgucma4756]

Great ! Im waiting for motor soft starter description!

[@gregoriglc]

Really good study of the transient state in the inverter. Good design for the relay timer as well. 10/10!!

[@johannes5219]

Could you also put an inductor in series? Would be more inefficient I think.. :/

[@caffeinatedinsanity2324]

If you don't forget to put a freewheeling diode, maybe. But don't forget you want as little power losses as possible, and coils... Well... Have a good amount of ESR. Not to mention coils can be expensive sometimes.

[@hadrast]

What about an inductor with an appropriately sized resistor (say 2Ω) in parallel though? The inductor could be a lot smaller (as in less inductive), since the resistor would bypass it and start charging the capacitor immediately, and the inductor windings could therefore be thicker to improve efficiency. Once the inductor starts passing current, the resistor would naturally drop out of the equation. Even better, the resistor would always be there to act as a drain for the inductor and tamp out potential oscillations.

[@johannes5219]

Ah okay, I see.. tanks for your answers :-)

[@fifaham]

Inrush current can cause Jitters or might confuse the digital circuit leading to severe outcome if used in medical or military applications. Caps, TVS, fly-back diodes and the like and the like can do magic in recovering from those problems.

[@cxmmax4265]

softstarter can be find in high end inverter, but it's cool anyway to see how to fix that

[@hydro_johnny]

@Rocky Robinson Because a capacitor this size probably costs more than this simple contraption and the inrush of that capacitor would be gigantic if ever discharged

[@cxmmax4265]

@Rocky Robinson it should work but you need an capacitor wich has no polarity, so if you attach a bigger load it wont work anymore, cause those capacxitor have a small capacity

[@TheRealMonnie]

I did a similar thing in a product using a comparator with an RC network, driving a low side n channel FET, which drove a high side p channel FET, so that after 0.5 seconds it shorted out a 24 ohm resistor. It worked great.

[@eritronc]

I think is better to use an p-channel mosfet, is more power saving...

[@yiqiwang4506]

Put an inductor in series with the input and an anti parallel diode with the inductor will solve the problem so much easier.

[@Videohead-eq5cy]

I was not paying attention till I heard "my final solution"

[@anotheruser9876]

And some say Germans don't have a sense of humour.

[@gunnaralv]

Dark...

[@JavaMac08]

Came here to see if anyone else noticed 🤣

[@frankescu]

And for my final solution... came looking for this

[@artbyrobot1]

he wasn't talking about jews when he said final solution

[@matthewmiller6068]

Interesting idea - I did something very similar to soft-start a car audio capacitor to stabilize my ham radio power supply. I just used the relay pull-in voltage to regulate when it turned on the relay though, seemed to be "good enough" and eliminate the majority of the spike.

[@bashaaksema94]

Cool project but why not use a second mosfet instead of the relay?

[@victorvillalba7401]

Maybe he didn't have a mosfet big enough?

[@alexoja2918]

All the current would have to go through the mosfet and stress it, requiring power components ("bigger mosfet") and producing heat. The problems this presents are greater than the required relay current.

[@johndavis6482]

I added a homade super cap to my solar system in line with battery and I had no problems with a pool pump draw ,without it the system All ways tripped saftey

[@johndavis6482]

The bms always pulls a load and the pump is on a timer for maxim use of sun

[@moiquiregardevideo]

@@alexoja2918 Obviously, the pump have much more current rush at start than the capacitor. Your system was already designed to limit in rush current. You didn't need that capacitor.

[@innovation4u336]

You are best teacher in the world. I always wait for your next video. Thanks man.

[@greatscottlab]

You're welcome :-)

[@vega1287]

"the post apocalyptic inventor" did so ething similar but he used a 555 timer , but hus cirquit was also made for mains voltage

[@vega1287]

@Michael Francis kzbin.info/www/bejne/laGonqOsmLOgjsk here you go you fond thr schematic at 2:12

[@Ben-yy7io]

I was just having a dilemma due to inrush current. Thank you for providing a possible solution.

[@elninorosario]

Can we just switch on that switch based on the voltage of the capacitor? .. (using a comparator maybe) .. and avoiding the delay circuit entirely!

[@tunahankaratay1523]

Surely, but it is a bit less convinient imo.

[@elninorosario]

@@tunahankaratay1523 or maybe just a constant current source using a mosfet?

[@tunahankaratay1523]

@@elninorosario Aww, that one would probably be problematic. It works like a very high frequency oscillator so it is not pure dc. You would need to smooth the current with an inductor but then you could also directly use an inductor. The inductor size will decrase with the circuit of course but it would still be very bulky and expensive. Relay circuits are even more efficient with big loads and it is super cheap and simple too.

[@tunahankaratay1523]

I think I need to clarify a bit more. The constant current circuit is a comparator with variable voltage input on one side and a current sensing resistor connected on the other input. The resistor connects to ground and the load. Since the mosfet gate has a capacitance in the picofarad range, the circuit acts like an RC oscillator. This oscillation is a square wave with changing duty cycle and frequency. It works at the end but the voltage isn't perfect dc. And the mosfet would waste more power than 0.5w(0.05 ohm*10 amp=5w, crazy)(the relay res is as negligible as the wire res).

[@robinbuster1323]

I really liked the thermistor trick, but your criticism was very fair.

[@johnemanuel5593]

Is that possible by simple inductor be used bcoz Sudden increase in current can be slowed down by inductor.... Plz answer

[@SelectLOL]

good question!

[@alexoja2918]

In some limited cases this can be done for low currents, but it's more akin to filtering and is expensive. Most important of all: an inductor usually produces inrush currents. That's why high inductive loads cause your lights to dim for a split second. Coils take current to energize.

[@SelectLOL]

@@alexoja2918 Yeah but in the sense of price i bet its cheaper then the realy and mosfet together.

[@viermidebutura]

short answer: yes long answer: the required inductor has to handle big currents and has to have a lot of inductance (in the H range) which means it will weight many kilos and cost a fortune

[@alexoja2918]

@@SelectLOL It would be a huge pile of copper and ferrite core, compared to a couple of generic components! The main efficient uses for transformers (dual inductors) are AC power transmission and SMPS circuits using high frequencies, and even in as small a use as the latter, they are usually among the most expensive components in the BOM.

[@Zebra_Paw]

I needed this for my new gaming audio setup, not because of the inrush but because it makes a huge noise in the speaker.

[@jamest.5001]

Why connect the inverter to the charge controller? That is not the best connection, most are connected directly to the battery! The output if the controller is usually designed for a street light, or night light, any reason for this?

[@gkdotclass]

MPPT charge controller has a battery protection feature which disconnects the load once an unsafe low battery voltage is reached . Also it includes load monitoring via the serial port

[@Phantomthecat]

Nice video, and good solution but I did the same thing in an easier way - I just connected my inverter up to the battery’s directly (through its own fuse) and don’t run it through the charge controller. Sure you won’t be able to monitor the load with the charge controller app, but it saves all of these issues. I don’t use the inverter much anyway and run almost everything I need to from 12 volts (lights, phone and laptop chargers, etc), but it is an easy fix. 👍. Keep up the great work, love your videos.

[@josephmagniez9580]

Nice video, thanks:) It would be interesting comparing this solution with another that uses a MOSFET as the main switch to reduce even more the power consumption of the circuit. With a 12V DC link it shall not be a huge challenge for you ;)

[@SianaGearz]

Inrush thermistor is actually quite adequate. For one it's more effective than it seems on repeat turn on, because at the core it's much hotter than at the surface, and cools down there much faster. For other the capacitors likely aren't discharged yet and so less inrush current limiting is needed. Also a choke can make for some additional inrush limiting.

[@Alex0474]

When you’re trying to make an arc lighter using advice from a GreatScott video but get a notification about a new video so you just go watch that one instead...

[@reedreamer9518]

Nice - that's a elegant solution for a soft starter. I used a 555 timer in my home built power supply project 6 months ago. But I like your MOSFET with the RC network better.

[@rockyrivermushrooms529]

I would like to see you make a VFD!

[@amirouw]

youre a great mentor .... greate Scott ❤

[@tunnelsnake627]

6:30 great Scott has Germanys final solution.

[@speedsterh]

Give the germans a break ! It must be tiresome for people like Scott to be constantly reminded of WWII

[@drmosfet]

In the old days, there was a special large mechanical relay that had two or more poles, and it's contacts throw could be adjusted independently. The first contact to make a contection had the current limiting resistor, the second contact would them bypass the resistor.